| 00:29.45 | Mirell | Mmm. home. |
| 00:41.25 | *** join/#edev veryevil12345 (n=so@client-82-2-88-36.manc.adsl.virgin.net) |
| 00:43.02 | Mirell | For those who want to feel old: http://www.beloit.edu/~pubaff/mindset/2011.php |
| 00:44.46 | *** mode/#edev [+o prpplague] by ChanServ |
| 00:45.00 | prpplague | well this is interesting |
| 00:45.17 | Cajaka | ? |
| 00:45.28 | prpplague | what do you make of a large order of hammer's from google? |
| 00:45.38 | SpeedEvil | 10, 100000? |
| 00:45.44 | prpplague | 25 |
| 00:45.50 | ds2 | OpenMoko phones in short supply? |
| 00:45.57 | SpeedEvil | 'that looks cool, let's get some for the office' |
| 00:46.00 | SpeedEvil | ds2: they're not |
| 00:46.12 | ds2 | cheaper then openmoko |
| 00:46.18 | SpeedEvil | ds2: not really |
| 00:46.25 | Cajaka | AML have a stock? |
| 00:46.30 | ds2 | aren't openmoko's $400ish? |
| 00:46.37 | SpeedEvil | ds2: $300 |
| 00:46.42 | ds2 | ah |
| 00:46.46 | ds2 | so bareboards are cheaper |
| 00:46.57 | SpeedEvil | ds2: and once you add a touchscreen, GPS, GSM, battery, ... |
| 00:47.41 | Mirell | Interesting. |
| 00:48.59 | SpeedEvil | You can make a small proto-phone with hammer |
| 00:49.06 | ds2 | how about JTAG is more accessible then on a openmoko? |
| 00:49.08 | SpeedEvil | and I've even got some sketches and preliminary BOMs |
| 00:49.19 | SpeedEvil | ds2: true - the connector on the hammer is nicer |
| 00:49.35 | ds2 | 1/3 isn't too bad ;) |
| 00:49.46 | SpeedEvil | ds2: 1/3? |
| 00:50.17 | SpeedEvil | And I feel at this point I should nitpick that neo1973 is the hardware, OM is an OS. |
| 00:50.21 | ds2 | 1/3 guessed reasons as to why :) |
| 00:51.57 | SpeedEvil | Internal stubby plugin to the debug port, with 'stuff' |
| 00:52.17 | ds2 | Neo1973 is just expensive at the moment...openezx hardware are half the price |
| 00:53.05 | SpeedEvil | You're missing the point somewhat. |
| 00:53.31 | SpeedEvil | openezx hardware is absolutely not open. |
| 00:53.42 | ds2 | :) |
| 00:54.11 | SpeedEvil | They do not want or support you flashing other OSs |
| 00:54.36 | SpeedEvil | The goal of the OM project is to commoditise phones for their parent company |
| 00:54.47 | SpeedEvil | So they can be on the PC model, just selling hardware to end users |
| 00:55.06 | SpeedEvil | rather than the lowest bidder for a Nokia 6666. |
| 00:55.24 | SpeedEvil | Because selling to users is where the profit is. |
| 00:55.34 | ds2 | and you are completely right; but for a platform to prototype a phone, openezx will suffice for |
| 00:55.59 | SpeedEvil | ds2: the neo has actual documented debug ports and solder-tos |
| 00:56.06 | ds2 | it is odd that google is even interested in a ARMv4 |
| 00:56.09 | SpeedEvil | Even supported. |
| 00:56.15 | ds2 | SpeedEvil: I am aware of that |
| 00:57.43 | landley | prpplague: hopefully to evaluate them and then buy LOTS more? :) |
| 00:58.22 | *** part/#edev veryevil12345 (n=so@client-82-2-88-36.manc.adsl.virgin.net) |
| 00:58.33 | prpplague | landley: indeed |
| 00:58.46 | prpplague | landley: looks like they were for the ireland offices |
| 00:59.12 | SpeedEvil | Also - I find the motorola phones listed on that page only marginally cheaper than the neo - unsubsidised. |
| 00:59.59 | prpplague | hmm |
| 01:01.20 | SpeedEvil | Send them back a personalised email? 'Thanks for your interest, If you buy our company, you get a 10% discount!' :) |
| 01:01.34 | prpplague | SpeedEvil: yea the boss is handling that |
| 01:01.51 | prpplague | but i'm seeing something interesting |
| 01:02.12 | SpeedEvil | It could simply be petty dev cash. |
| 01:02.38 | SpeedEvil | But I hope it's more. |
| 01:02.43 | roxfan | people, what are typical voltages on jtag pins? |
| 01:02.54 | SpeedEvil | roxfan: what the voltage is on the circuits VCC |
| 01:03.01 | SpeedEvil | roxfan: 5 or 3.3V |
| 01:03.04 | roxfan | on all of them? |
| 01:03.26 | SpeedEvil | roxfan: no - you mean you want to find jtag pins on an unknown thingy? |
| 01:03.31 | roxfan | yes |
| 01:03.42 | ds2 | SpeedEvil: they can be had brand new for $139.99 |
| 01:03.51 | ds2 | so it is more the marnginally cheaper |
| 01:03.53 | SpeedEvil | ds2: unsubsidised? |
| 01:04.00 | ds2 | SpeedEvil: yep. unlocked |
| 01:04.08 | SpeedEvil | ds2: where? |
| 01:04.14 | roxfan | here's what i have 1) 3.12 2) 2.65 3) 3.22 4) 3.22 5) 0.26 6) 0.08 7) 3.10 (resets) |
| 01:04.17 | SpeedEvil | ds2: and which phone is that |
| 01:04.20 | roxfan | and 8-14=gnd |
| 01:04.29 | ds2 | SpeedEvil: Amazon; A780 |
| 01:04.47 | ds2 | the one that has been most worked on by the ezx folks |
| 01:04.50 | roxfan | it's a pxa270 btw |
| 01:04.54 | ds2 | so at less then 1/2 the price.... |
| 01:07.08 | SpeedEvil | ds2: somewhat unfair to compare a dev platform shipping in teh 1000s to a production one selling in the millions. I'm unsure if those are legally or illegally unlocked - I note the MSRP is $499 |
| 01:07.41 | ds2 | SpeedEvil: *nod* and the A780 is really at the end of its life; I suspect those are clearance units |
| 01:08.11 | ds2 | SpeedEvil: still, for experimenting, that price is right. a GSM module cost more then that! |
| 01:08.34 | *** join/#edev redblue (i=star@ppp042.110-253-207.mtl.mt.videotron.ca) |
| 01:08.36 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 01:08.37 | *** join/#edev agaffney (n=agaffney@gentoo/developer/pdpc.active.agaffney) |
| 01:08.47 | SpeedEvil | ds2: actually not. I got one for $40 |
| 01:08.53 | SpeedEvil | ebay china ++ |
| 01:08.57 | ds2 | heh |
| 01:09.06 | ds2 | I do not do ebay so that is a irrelevant market |
| 01:09.17 | SpeedEvil | even sparkfun only want $79 IIRC |
| 01:09.30 | ds2 | $79!?! |
| 01:09.47 | SpeedEvil | Maybe I misremembered. |
| 01:09.58 | ds2 | cheapest one I see is $149.95 |
| 01:10.01 | ds2 | oops 145.95 |
| 01:10.56 | ds2 | SpeedEvil: come to think of it, I think SFE was clearancing out a few returned units last year for around that price |
| 01:11.13 | RobotGuy | prpplague: Are you here? |
| 01:12.16 | SpeedEvil | Mine was a triband benq one, with even a free protoboard |
| 01:23.42 | RobotGuy | I just bought a WiFi-G router and USB WiFi-G adapter. |
| 01:29.53 | SpeedEvil | So? |
| 01:36.19 | RobotGuy | It was a pretty good price for both - $111.00 |
| 01:37.59 | SpeedEvil | Not very terrible. |
| 01:40.32 | prpplague | sorry phone |
| 01:42.42 | RobotGuy | I didn't have to pay shipping. |
| 01:42.51 | RobotGuy | And I spent my money locally. |
| 01:43.03 | prpplague | landley: interesting |
| 01:43.19 | prpplague | landley: the order for the hammer boards was marked for the "L'Ève future Department" |
| 01:43.28 | RobotGuy | prpplague: What are the exact build options to get C and C++ support? |
| 01:43.35 | *** join/#edev _lifeless (n=patrick@90.150.124.113) |
| 01:43.55 | prpplague | RobotGuy: just need to turn on the c++ in the build config |
| 01:44.18 | prpplague | landley: i did a google on L'Eve Future |
| 01:44.30 | prpplague | landley: took me to this page - http://en.wikipedia.org/wiki/Android |
| 01:44.41 | RobotGuy | prpplague: No shared libgcc? |
| 01:44.41 | prpplague | coincidence? |
| 01:44.50 | prpplague | RobotGuy: yea that should be on by default |
| 01:45.02 | SpeedEvil | I have a vagur recollection that that's their general research arm |
| 01:45.27 | RobotGuy | prpplague: I couldn't get a toolchain with those options all on to build. |
| 01:46.33 | RobotGuy | prpplague: Should I have C++ and libstdc++ turned on? |
| 01:50.32 | prpplague | RobotGuy: thats the same config that's in the prpplague home dir |
| 01:50.37 | prpplague | RobotGuy: built fine for me |
| 01:50.39 | RobotGuy | I have both C++ options turned on |
| 01:51.14 | prpplague | RobotGuy: what failes |
| 01:51.19 | prpplague | RobotGuy: ? |
| 01:53.26 | RobotGuy | prpplague: I'll let you know when this toolchain build bombs. |
| 01:54.46 | RobotGuy | prpplague: On the Gadget Ethernet, I can ping in both directions, but can't get beyond intrepid. I think it has to be a routing issue, or maybe my kernel does not have some options it needs. |
| 01:55.02 | prpplague | thats routing |
| 01:55.22 | prpplague | you need to add the default gw and also set ip forwarding |
| 01:56.00 | RobotGuy | prpplague: I have the gateway set to eth0 on intrepid at present. |
| 01:56.16 | RobotGuy | prpplague: The static IP address of eth0 |
| 01:56.45 | RobotGuy | prpplague: I may not have forwarding turned on now. |
| 01:57.13 | SpeedEvil | RobotGuy: have you looked at the FAQ I pointed you to on the OM wiki? |
| 01:57.29 | SpeedEvil | RobotGuy: on how to get forwarding working on your desktop/laptop side |
| 01:57.29 | prpplague | RobotGuy: no, you need to set the default gw on the hammer |
| 01:57.52 | RobotGuy | SpeedEvil: I looked at some of it and added some stuff to my config. |
| 01:57.56 | RobotGuy | prpplague: Yes, I know. |
| 01:58.07 | SpeedEvil | RobotGuy: do you now have iptables masq installed? |
| 01:58.27 | SpeedEvil | RobotGuy: and done the echo >/proc/... thing? |
| 01:58.55 | RobotGuy | SpeedEvil: Yes. I have forwarding turned on. |
| 01:59.17 | SpeedEvil | does ping an.other.ip work? |
| 01:59.38 | SpeedEvil | you will of course need to route add default gw your.laptop.ip |
| 02:00.03 | RobotGuy | SpeedEvil: As I said earlier, I can ping in both directions - Hammer to intrepid and intrepid to Hammer. Just nothing goes out from intrepid. |
| 02:00.24 | RobotGuy | Default gateways are added. |
| 02:00.39 | SpeedEvil | RobotGuy: do the iptables lines run without error on your desktop/laptop? |
| 02:00.50 | RobotGuy | The Hammer default gateway may not be right though. |
| 02:02.06 | prpplague | RobotGuy: what did you add on the hammer was the default gw? |
| 02:03.20 | *** join/#edev RobotGuy (n=robotguy@dsl093-038-072.pdx1.dsl.speakeasy.net) |
| 02:05.16 | prpplague | RobotGuy: what did you add on the hammer was the default gw? |
| 02:05.26 | RobotGuy | prpplague: The IP of intrepid's eth0 |
| 02:05.36 | prpplague | RobotGuy: wrong up then |
| 02:05.54 | prpplague | RobotGuy: default gw should be the ip address of intrepids usb0 interface |
| 02:06.11 | RobotGuy | prpplague: OK, easy to change. |
| 02:07.15 | SpeedEvil | prpplague: actually not |
| 02:07.16 | SpeedEvil | route add 10.0.0.66 usb0 |
| 02:07.16 | SpeedEvil | root@fic-gta01:~$ route add default gw 10.0.0.66 |
| 02:07.34 | SpeedEvil | and then I can ping google - after removing my original default IF |
| 02:07.42 | SpeedEvil | Or at least - not in all cases |
| 02:07.59 | SpeedEvil | It happens to work in mine - that's my ethernet IP (well - eth0:1) |
| 02:08.11 | prpplague | i usually add the ip address of the remote pc, even on dialup pppd connections |
| 02:08.34 | SpeedEvil | I'm fairly sure 'modern' ifconfigs'll do that for you |
| 02:08.34 | RobotGuy | prpplague: I did "route add default 192.168.50.2 usb0" |
| 02:09.03 | SpeedEvil | should be route add gw 192.... |
| 02:09.13 | SpeedEvil | I don't know if the two combined will work |
| 02:09.20 | ds2 | think you are replying on proxy arp routing |
| 02:09.30 | prpplague | bbiab |
| 02:09.55 | SpeedEvil | RobotGuy: try route add default gw 192.168.50.2 |
| 02:10.03 | SpeedEvil | (if you can ping 192.168.50.2) |
| 02:10.04 | RobotGuy | SpeedEvil: I did that |
| 02:10.31 | RobotGuy | I forgot to add the 'gw' to the above |
| 02:11.19 | SpeedEvil | If you've done that, and can ping 192.168.50.2, and now can't ping an external site by IP, you've screwed up somewhere in the host-side routing |
| 02:12.09 | RobotGuy | SpeedEvil: I figured that out already. |
| 02:12.24 | RobotGuy | I said it's probably a host config problem |
| 02:12.41 | SpeedEvil | RobotGuy: do you have a spare PC on the network you can try it with? |
| 02:13.50 | RobotGuy | I can not ping anything outside of intrepid. |
| 02:13.54 | RobotGuy | From hammer |
| 02:14.23 | SpeedEvil | I mean so you can try getting that other PC gatewayed through intrepid |
| 02:14.56 | RobotGuy | I don't want that other PC gatewayed through intrepid. |
| 02:15.29 | RobotGuy | That configuration is not doable because I don't have another ethernet card in intrepid. |
| 02:17.01 | SpeedEvil | RobotGuy: does iptables -A POSTROUTING -t nat -j MASQUERADE -s 192.168.0.0/24 and echo 1 > /proc/sys/net/ipv4/ip_forward and iptables -P FORWARD ACCEPT all complete without error? |
| 02:17.12 | SpeedEvil | (host side) |
| 02:19.41 | RobotGuy | SpeedEvil: All those complete with no error. |
| 02:20.20 | RobotGuy | I have a static IP on intrepid's eth0 |
| 02:21.23 | RobotGuy | I think it could be a kernel config error. I never really trust stock kernels. |
| 02:30.26 | ds2 | RobotGuy: what are you trying to ping that you can't get to? |
| 02:31.06 | SpeedEvil | RobotGuy: what does iptables -v -L show? |
| 02:31.10 | SpeedEvil | (pastebin) |
| 02:31.14 | RobotGuy | Anything outside of intrepid (host to hammer). |
| 02:32.36 | RobotGuy | SpeedEvil: http://www.pastebin.ca/888407 |
| 02:33.23 | ds2 | RobotGuy: any of them Linux hosts with tcpdump installed? |
| 02:33.54 | RobotGuy | intrepid has tcpdump |
| 02:34.54 | ds2 | something beyond it so you can confirm if the masq stuff is working |
| 02:35.33 | RobotGuy | ds2: NOTHING BEYOND intrepid from Hammer. |
| 02:35.46 | ds2 | didn't you say you can ping intrepid? |
| 02:35.52 | RobotGuy | Yes |
| 02:39.59 | RobotGuy | prpplague: I just built a new toolchain with the options you said all set. I still can't build libusb. |
| 02:41.11 | SpeedEvil | RobotGuy: oops - missed that - sorry - arging with a silly person |
| 02:45.16 | SpeedEvil | RobotGuy: iptables -t nat -L |
| 02:45.31 | SpeedEvil | Chain POSTROUTING (policy ACCEPT) |
| 02:45.31 | SpeedEvil | target prot opt source destination |
| 02:45.31 | SpeedEvil | MASQUERADE all -- 192.168.0.0/24 anywhere |
| 02:45.36 | SpeedEvil | should be one of the outputs |
| 02:47.08 | RobotGuy | SpeedEvil: http://www.pastebin.ca/888419 |
| 02:48.01 | SpeedEvil | Ok - that differs from mine - not the prot all |
| 02:48.29 | RobotGuy | This sort of routing and such is new to me, so I am learning it bit by bit. |
| 02:49.30 | SpeedEvil | I have iptable_nat ip_tables nf_nat x_tables nf_conntrack_ipv4 nf_conntrack modules |
| 02:49.32 | SpeedEvil | installed |
| 02:50.02 | SpeedEvil | which mention nat in lsmod|grep nat |
| 02:51.13 | RobotGuy | SpeedEvil: http://www.pastebin.ca/888424 |
| 02:56.35 | landley | prpplague: yeah, the google android project is the biggest likely user of arm. :) |
| 02:59.33 | RobotGuy | landley: Hello |
| 02:59.54 | SpeedEvil | RobotGuy: sorry - I'd suggest going through the networking/masq section of make menuconfig, and ensure you've got everything selected. Also check for a recent version of |
| 03:00.03 | SpeedEvil | RobotGuy: iptables (though that shouldn't be it) |
| 03:00.42 | RobotGuy | SpeedEvil: I'm going to try configuring a 2.6.24 kernel for intrepid. |
| 03:02.36 | SpeedEvil | RobotGuy: as a stupid starting point - |
| 03:02.46 | SpeedEvil | http://www.mauve.plus.com/config.gz |
| 03:02.49 | SpeedEvil | is my current config |
| 03:03.21 | SpeedEvil | err - ignore that |
| 03:03.51 | SpeedEvil | http://www.mauve.plus.com/config.gz |
| 03:03.53 | SpeedEvil | is my current config |
| 03:03.55 | SpeedEvil | (really) |
| 03:09.04 | ds2 | RobotGuy: still not working? |
| 03:09.15 | RobotGuy | ds2: Not yet |
| 03:09.20 | RobotGuy | Nothing has changed. |
| 03:09.43 | ds2 | RobotGuy: mind if I ask you some questions you may have answered? (I am waiting for a build so...) |
| 03:10.27 | RobotGuy | ds2: I don't want to repeat what I have already told SpeedEvil. Otherwise, sure. |
| 03:10.56 | ds2 | RobotGuy: just basic config - what is the IP you are using on the hammer's usb0? |
| 03:11.09 | SpeedEvil | ds2: he can ping abck and forth |
| 03:11.15 | ds2 | trying to draw a network diagram |
| 03:11.35 | SpeedEvil | ds2: the problem is the routing on the host side - it's not forwarding |
| 03:11.39 | RobotGuy | ds2: SpeedEvil and I have already covered basic configuration. |
| 03:11.52 | RobotGuy | SpeedEvil: Yep, as I thought it would be. |
| 03:12.31 | ds2 | FINE |
| 03:12.46 | ds2 | solve it yourself |
| 03:13.19 | SpeedEvil | ds2: most of the simple stuff is covered in backscroll |
| 03:13.21 | RobotGuy | ds2: There is no need to get pissy ds2. |
| 04:03.15 | *** join/#edev linac (n=lin@121.41.15.117) |
| 04:12.28 | *** join/#edev Dr{Wh0} (n=mathewss@dev.null.nutech.com) |
| 04:16.10 | *** join/#edev ohsix (i=ohsix@66.220.111.222) |
| 04:23.04 | *** join/#edev GPSFan (n=kenm@12.10.255.246) |
| 04:23.05 | *** mode/#edev [+o GPSFan] by ChanServ |
| 04:26.20 | *** join/#edev dijenerate (n=dijenera@72.51.115.81) |
| 04:28.31 | *** join/#edev mauve_ (n=mauve@212.159.7.165) |
| 05:28.34 | *** join/#edev DrZimmerman (n=theo@tbaumgartner.broker.freenet6.net) |
| 07:48.43 | *** join/#edev __lifeless (n=patrick@90.150.221.156) |
| 08:24.20 | *** join/#edev emil|k (n=emil@e179234066.adsl.alicedsl.de) |
| 09:01.15 | *** join/#edev redblue (i=star@ppp019.108-253-207.mtl.mt.videotron.ca) |
| 09:18.09 | *** join/#edev _lifeless (n=patrick@90.150.114.125) |
| 09:28.58 | *** join/#edev lyakh (n=lyakh@p57BD313F.dip0.t-ipconnect.de) |
| 09:38.23 | *** join/#edev __lifeless (n=patrick@89.20.123.195) |
| 10:50.46 | *** join/#edev steliosk_ (n=Stelios@athedsl-317459.home.otenet.gr) |
| 11:56.30 | *** join/#edev roxfan (n=dunno@30.49-136-217.adsl-dyn.isp.belgacom.be) |
| 13:03.39 | *** join/#edev linac (n=lin@121.41.15.117) |
| 13:47.51 | *** join/#edev bisy (n=bisy@host86-152-80-191.range86-152.btcentralplus.com) |
| 13:48.12 | *** join/#edev emil|k (n=emil@f054013167.adsl.alicedsl.de) |
| 14:00.39 | *** join/#edev ico2 (n=ico2@modem-2730.cheetah.dialup.pol.co.uk) |
| 14:27.51 | *** join/#edev GPSFan (n=kenm@12.10.255.246) |
| 14:27.51 | *** mode/#edev [+o GPSFan] by ChanServ |
| 14:35.09 | *** join/#edev prpplague (n=dave@ppp-70-244-166-105.dsl.rcsntx.swbell.net) |
| 14:38.21 | prpplague | ho ho hum |
| 14:38.38 | prpplague | well.... ol dave hylands finally got squeak working properly |
| 14:38.57 | *** mode/#edev [+o prpplague] by ChanServ |
| 14:40.28 | GPSFan | prpplague: what was with the fp errors in the test suite? |
| 14:40.58 | *** join/#edev ico2 (n=ico2@modem-2730.cheetah.dialup.pol.co.uk) |
| 14:41.13 | prpplague | GPSFan: well for the the buildroot portion i solved all of those, it was mainly just compile options |
| 14:41.36 | prpplague | GPSFan: as far as squeak fp, it was config option, it was hardcoded for xscale support |
| 14:41.55 | prpplague | GPSFan: rather than have an option for different endian configurations |
| 14:42.03 | GPSFan | that's good, now that it "works" what can one do with it? |
| 14:42.47 | prpplague | GPSFan: well as far squeak goes, i have no clue for me, but for the robotics and space/near-space developers, it seems to be the environment of choice |
| 14:43.18 | prpplague | GPSFan: did you see we got a big order from google? |
| 14:43.38 | GPSFan | hmmm, well maybe google is going to use it as a dev platform for "summer of code" work |
| 14:44.04 | *** join/#edev g1powermac (n=g1powerm@unaffiliated/g1powermac) |
| 14:44.10 | prpplague | GPSFan: yea |
| 14:44.11 | GPSFan | as landley wanted with the nail, just take it with you to the coffee shop and have at it. |
| 14:44.44 | prpplague | GPSFan: it was marked for "L'Eve Future Department" in dublin, ireland |
| 14:45.26 | prpplague | GPSFan: if you google(how funny is that) L'Eve Future, you get a wikipedia page for the first usage of the word Android |
| 14:46.00 | SpeedEvil | I think it's their 'wacky bluesky ideas' depy |
| 14:46.04 | GPSFan | hmm google + android + hammer =??? |
| 14:46.28 | prpplague | no clue, just thought it was funny |
| 14:46.43 | SpeedEvil | I can't imagine it'd make sense to use a modular CPU in a phone |
| 14:47.03 | prpplague | me either |
| 14:47.04 | SpeedEvil | I assume the BOM cost of the hammer in 10Ks'd be $30 or so? |
| 14:47.22 | prpplague | SpeedEvil: closer to $20 in 10k |
| 14:47.59 | SpeedEvil | I suppose it's pointless to speculate :) |
| 14:48.13 | prpplague | indeed |
| 14:48.14 | GPSFan | pointless, but fun. |
| 14:48.32 | prpplague | the boss sent some emails to the contact late last night |
| 14:48.38 | prpplague | maybe we'll here something back |
| 14:49.28 | prpplague | SpeedEvil: really if you did a 20k build you could do all CoB and get the cost below $10 |
| 14:50.42 | prpplague | http://www.informationweek.com/blog/main/archives/2007/11/google_guys_tal.html;jsessionid=4ZTDC11IO01SCQSNDLPSKH0CJUNN2JVN |
| 14:52.00 | prpplague | GPSFan: anyway, guess what? i got tasked on friday to upgrade a legacy product to a color tft display |
| 14:52.03 | prpplague | GPSFan: sooooo |
| 14:52.18 | prpplague | GPSFan: i'm gonna use the hammer to do all my testing |
| 14:52.30 | prpplague | GPSFan: hehe lots of samples :) |
| 14:52.39 | GPSFan | prpplague: wasn't that the original plan? |
| 14:52.54 | GPSFan | what size tft? |
| 14:52.58 | prpplague | GPSFan: sorry, what was the plan? |
| 14:53.01 | prpplague | GPSFan: 2.8" |
| 14:53.36 | prpplague | GPSFan: similiar platform - http://www.elinux.org/M5900 |
| 14:54.48 | prpplague | is 8:57am too early to have a beer? |
| 14:55.12 | GPSFan | prpplague: well the original plan was to make the hammer attractive for all sorts of projects, seems like you are doing just that. |
| 14:55.44 | prpplague | GPSFan: indeed, thats what we've been working on |
| 14:56.13 | prpplague | GPSFan: the thing about the lcd stuff though is that its for AML, so i get standard "work day" time to spend on it |
| 14:56.31 | prpplague | GPSFan: so if i do my testing with the hammer, and get lots of samples, i can document them on the wiki |
| 14:56.36 | prpplague | two birds with one stone |
| 14:56.59 | GPSFan | beer at 8:57 might have some nasty side effects, like more beer at 9:10, 9:30, 9: 50 |
| 14:58.41 | prpplague | hmm, well its 4:57pm in capetown south africa, so its 5 somewhere, |
| 14:58.47 | prpplague | guess that makes it ok |
| 15:00.14 | prpplague | hehe |
| 15:01.40 | GPSFan | prpplague: hey afterall it's superbowlsunday tomorrow, so you gotta get in practice/shape. |
| 15:02.09 | prpplague | GPSFan: hehe, indeed that would be a valid argument if i actually watch the superbowl |
| 15:02.25 | GPSFan | yeah, me too. |
| 15:04.18 | prpplague | GPSFan: z2 dev? |
| 15:04.56 | prpplague | GPSFan: i'm trying to stay unplugged per se this weekend, got a long week coming up |
| 15:05.54 | GPSFan | yeah, I'm trying to get wavemon working. worked great (my version) on the Z1, and most of it works on the Z2, but when it scans for acces points it segfaults. It may have to do with inline code which has had some breakage in gcc 4.x.x. |
| 15:06.46 | prpplague | ahh |
| 15:06.52 | GPSFan | iwlist eth0 scan works fine on the z2, and it uses almost the exact same code ( I ripped it out of wireless tools) to do the scanning. |
| 15:06.57 | prpplague | they find jtag points yet? |
| 15:08.27 | GPSFan | T0mW needs to wipe all the parts off his second Z2, the folks at zipit say that jtag wasn't brought out the rear connector, and are vary closed mouth about whether it (or the 2 serial ports) are accessable on the board. |
| 15:09.25 | GPSFan | I can't imagine an Xscale design that doesn't have jtag for production testing and initial flash programming. |
| 15:09.58 | prpplague | indeed |
| 15:10.33 | prpplague | i guess it seems the zipit folks still haven't learned their lesson |
| 15:11.02 | GPSFan | they use blob for a bootloader, and blob really likes to use the serial port, it's even set up in the blob patches they released, and the kernel command line has the console=ttyS??? |
| 15:12.02 | GPSFan | ttyS2,115200 |
| 15:12.12 | SpeedEvil | I suppose one option would be to do COB assembly |
| 15:12.34 | SpeedEvil | then JTAG program it, and cover the blobs once you've verifird |
| 15:13.17 | SpeedEvil | Though there are probably easier ways. |
| 15:13.34 | GPSFan | the Z2 board is mostly bga for the major parts, a few small outlint leaded parts. and the flash chip has a green dot (paint) on it. Usually indicates a pre-programmed part. |
| 15:16.06 | SpeedEvil | I suppose if you make it from BGA, bury the critical traces, solder the programmed flash on, and have a bootloader that wants a crypto-signed blob for upgrades over some port, you've covered most of the bases. |
| 15:16.50 | SpeedEvil | Yes, you can break it but it gets annoying and it's not easier on the tenth device than on the first, like many hacks |
| 15:17.31 | GPSFan | mostly, except for production testing, I wouldn't want to throw away too many repairable devices. |
| 15:18.26 | SpeedEvil | Bring out JTAG through something you can disable |
| 15:19.28 | GPSFan | or to some "test points" there are quite a few of those scattered about, nicely gold plated too. just need to figure out which ones go where. |
| 15:19.48 | SpeedEvil | I should ahve bookmarked that |
| 15:19.57 | SpeedEvil | I saw a microcontroller project |
| 15:20.06 | SpeedEvil | You connect it to a lot of pins |
| 15:20.17 | SpeedEvil | It figures out what ones are JTAG and what the pinout is |
| 15:20.23 | GPSFan | the ones I've looked at under a microscope have evidence of being probed. so they do have some sort of production tester. |
| 15:20.46 | GPSFan | I remember seeing that article. I didn't bookmark it either. |
| 15:21.06 | SpeedEvil | Are there any cheap bare board-netlist vendors? |
| 15:23.06 | GPSFan | you mean someone that takes a bare board and procuces a schematic/netlist.... I did it for most of the Z1, and tuxscreen, T0mW did a much better job for the Z1 (he completely wiped the board and traced out everything) |
| 15:23.18 | SpeedEvil | Yes. |
| 15:23.24 | SpeedEvil | Ideally automated |
| 15:24.19 | SpeedEvil | I suppose x-rays'd help |
| 15:24.37 | GPSFan | I've herad of some folks that do that sort of things for IC's so I'm sure that there are some that do it for PCB's. as for cheap, I don't think they know how to spell that word. |
| 15:25.08 | SpeedEvil | Probably not. |
| 15:26.39 | GPSFan | this guy did/does stuff like that http://wifi.openschemes.com/ see the "Serious dissassembly" section about half way down. |
| 15:27.50 | SpeedEvil | I really really hope what samsung does doesn't become more common. |
| 15:28.00 | SpeedEvil | (no datasheets) |
| 15:28.22 | SpeedEvil | Not only from the hacking, also from the screwing over the small developer perspective. |
| 15:31.27 | prpplague | the jtag diag device was posted on hackaday |
| 15:31.51 | prpplague | GPSFan: send me a board, we can x-ray it |
| 15:37.03 | prpplague | SpeedEvil: the datasheets are available, you just have to request them |
| 15:39.39 | SpeedEvil | prpplague: maybe I need to try again. |
| 15:39.44 | SpeedEvil | Diddn't work last time. |
| 15:39.57 | prpplague | SpeedEvil: what datasheet do you need? |
| 15:40.11 | *** join/#edev ico2 (n=ico2@modem-3907.chameleon.dialup.pol.co.uk) |
| 15:40.17 | GPSFan | prpplague: I only have one, T0mW has 2, he bought the second specifically to wipe it, but has been reluctant to sinece it will destroy the board. for the Z1 he and I got broken ones from fiolks that hosed up the serial port mod or broke the display. |
| 15:40.18 | SpeedEvil | I found it - google eventually |
| 15:40.33 | SpeedEvil | Looking up data on teh 2442 for freerunner |
| 15:40.43 | prpplague | SpeedEvil: should have asked me |
| 15:40.53 | prpplague | SpeedEvil: i have datasheets on all of the s3c24xx series |
| 15:41.18 | SpeedEvil | It popped right up on google. |
| 15:41.19 | prpplague | GPSFan: well, let me know |
| 15:41.38 | SpeedEvil | Thanks though. |
| 15:41.44 | SpeedEvil | Some won't |
| 15:42.25 | GPSFan | ok, thanks. |
| 15:50.05 | SpeedEvil | GPSFan: that is rather completely awesome |
| 16:05.15 | prpplague | well its time for a road trip :( |
| 16:05.22 | prpplague | bb this evening |
| 16:05.34 | prpplague | pooooooooof |
| 16:19.12 | *** join/#edev chouimat|away (n=dieu@r2351064.cidc.net) |
| 16:37.05 | GPSFan | SpeedEvil: yeah that sure is some "serious" dissassembly |
| 17:02.52 | *** join/#edev emil|k (n=emil@f054013167.adsl.alicedsl.de) |
| 17:40.18 | *** join/#edev dijenerate (n=dijenera@72.51.115.81) |
| 17:41.38 | *** join/#edev adyer (n=adyer@adsl-69-223-13-78.dsl.chcgil.ameritech.net) |
| 17:59.50 | *** join/#edev bbradley (n=bbradley@82-32-35-42.cable.ubr05.azte.blueyonder.co.uk) |
| 18:37.43 | *** join/#edev redblue (i=star@ppp105.108-253-207.mtl.mt.videotron.ca) |
| 18:50.33 | *** join/#edev DrZimmerman (n=theo@tbaumgartner.broker.freenet6.net) |
| 18:51.30 | *** join/#edev Redhatter (n=vk4fsjl@2001:388:f000:0:0:0:0:279) |
| 20:52.23 | *** join/#edev lifeless (n=patrick@89.20.117.72) |